March 2021 Roadmap Update, Part 2

In our last update, we released our most recent technical roadmap and outlined some of the major contours of the road ahead. As promised, “Roadmap Part 2” highlights a selection of major achievements from the past year.

As of last spring, we were testing a comparatively minimal version of Galileo, designed to become the front-end entry point to the planned Hypernet computing marketplace. Over the past year, Galileo has become more and more powerful with the addition of software integrations in target industries, the “Missions” project-based framework, “Cargo Bay” storage features, and Jupyter Notebooks, to name just a few key developments. Alongside our work towards facilitating microtransactions for metered services, these milestones are paving the way towards testnet, which we currently expect to reach by the end of Q2.

Technical Roadmap Towards Testnet. Click here to expand.

Some highlights:

Spring was an important time for Hypernet Labs. In the face of the historic challenge that was only beginning to come into view, we rose to the occasion and worked to fulfill our mission of supporting scientific discovery and progress. In the midst of crisis, we saw an opportunity to work towards fulfilling our mission of realizing ethical ubiquitous computing.

As such, we worked quickly to onboard scientists and researchers around the world, in various fields, as users of the Galileo web application, powered by Hypernet. This group included epidemiologists, biologists, chemists, and others who were tracking the virus or developing medications to treat the illness. To support these researchers, we worked to ensure that specialized software could be run smoothly and accelerated with the help of the Galileo app. We quickly found that pharmaceutical research greatly benefits from seamless access to additional power, as drug development research requires thousands of runs and iterations.

We understood, at this point, that further development of the app would greatly benefit from close work with test users to stress test the application for scalability, identify friction points, and help improve usability and decide on new features. By May, we had achieved our initial goal of onboarding a critical number of test users, reaching over 1000 signups for app testing.

MAGAZINE: Brazilian Researchers Use Galileo for COVID-19 Drug Repurposing Project

By summer we had begun to collect several fascinating case studies to showcase in our new Galileo Magazine. We were also acknowledged in Harvard Data Science Review and Computational Biology and Chemistry for our work with statisticians, epidemiologists and molecular dynamics researchers.

Anastasios Angelopoulos’ and Michael I. Jordan’s HDSR paper aimed to help policymakers better allocate COVID relief and health infrastructure resources across populations by providing a more accurate method for estimating relative case fatality ratios. We are proud to have contributed to this important study, which was also highlighted in VentureBeat at the end of May.

VentureBeat: Researchers Use Predictive Models to CalculateMore Accurate COVID-19 Case Fatality Ratios

In July we continued working to integrate software packages in target fields, including pharmaceutical research, biotech, economics, and civil engineering. For example, we completed our integration with FLO-2D flood modeling software and made major progress with other hydrology and hydraulics simulators.

MAGAZINE: Cleaning Up Lake Erie with Compute

Towards the end of summer we reached an important milestone with the release of Galileo 3.0. This was a significant technical achievement, which required foundational innovation and complex problem solving across the entire team. It also required extensive stress-testing with power users.

Galileo 3.0 introduces the Missions feature to the platform, which presents several benefits.

  • With Missions, users can avoid re-uploading large data files with every run. Instead, they can make changes to the simulation or analysis and run the project over and over without starting from scratch every time.
  • Missions also allows users to organize their work according to project type, without having to scroll through pages of individual, uncategorized runs.
MAGAZINE: Monash University Researchers Shift Focus to COVID Antivirals and Get Results with Galileo

At the end of summer, we began collaborations with Connext and 3box as we worked on Hypernet protocol development. We also continued to grow our pool of supported software in view of making Galileo into a near-universal platform.

In water resource management, we worked with users and software developers of FLO2D, PCSWMM, HEC-RAS, and other important platforms. Galileo helps these engineers to complete vastly more work in less time, better organize their simulations, and focus on their models without needing to learn to set up their own computing infrastructure.

In pharmaceuticals, we completed software integrations to support Gromacs, AutoDock Vina, Quantum Espresso, and others. Galileo helps pharmaceutical scientists to speed up their work, to run more detailed and accurate simulations, and to screen a greater number of drugs in a much shorter span of time.

In October we were very happy to announce that Hypernet Labs was accepted into StartX, Stanford’s prestigious and highly competitive startup accelerator program. We’ve been taking advantage of this exciting opportunity to better grow our company and further our mission since the program’s start this fall.

On the technical front, Galileo’s new Missions feature went to wider release, and we began beta testing our Jupyter Notebook-Galileo integration.

We also began collaborating with engineers and chemists at the Open Force Field Initiative (OpenFF) in their work on QCFractal, now in beta. The goal is the development of a tool for pharmaceutical research that could be distributed and run within Galileo.

Finally, we announced initial collaborative work with our first large compute provider to bring their High Performance Computing resources online and make them accessible in Galileo. We began to carry out initial technical tests. This is significant as a proof-of-concept for HPC-Galileo integration (necessary for certain kinds of supercomputing) as well as a mile marker on the road to realizing the eventual Hypernet marketplace for computing power.

At the start of winter, notable new features in Galileo included the integration of various storage options, or “cargo bays,” and an in-app text editor. We also continued to work to streamline the workflow for the Jupyter Notebook integration, maximizing ease and efficiency.

The hugely significant cargo bay feature allows users to connect to their own object-storage providers, including popular options like Dropbox and Box as well as competitive alternative options. Cargo bays make it even easier for users to work with large datasets without worrying about uploading data multiple times. With the text editor, users can now edit the text of the files directly in Galileo without having to delete and re-upload new versions.

In December we co-hosted a webinar with FLO-2D as a successful proof of concept for future educational co-marketing initiatives with partner software developers.

At the start of the new year, we showcased our work with partners in the blockchain space and unveiled the first preview demo of the Hypernet Core, the layer 2 scaling piece of the Hypernet Protocol stack.

Our work with Storj gives Galileo users increased flexibility in their data storage options, as they can now use Storj’s Tardigrade service as a cargo bay attached to their Galileo account.

We also announced our work with Connext, with the goal of more readily enabling the high-throughput payments required by a decentralized marketplace. The Hypernet Protocol aims to address blockchain throughput issues by incorporating Connext’s latest protocol, Vector, into a seamless experience for developers and end users.

We are moving full steam ahead to deploy the Hypernet Protocol to empower the community and support a broad range of real-world use cases. With the first application powered by the Protocol (Galileo) and the Hypernet Core workstreams coming together, we are gearing up to embark upon the first phases of the forthcoming public testnet.

In our next update, look for our first demo of the HyperConnector, which is designed to make it relatively easy for developers to enable their applications to consume the Hypernet Protocol.

Hypernet harnesses blockchain tech and cloud to power data science, simulations, AI, and other computing work.